Diastereomer Splitting in the 95Mo NMR Spectra of Compounds Differing only in the MO Configuration

(1)

Inorganica Chimica Acta, 96 (1985) LS-L9 LS

Diastereomer Splitting in the 95Mo NMR Spectra of Compounds Differing only in the MO Configuration

HENRI BRUNNER*, PETER BEIER, ERICH

FRAUENDORFER, MANFRED MUSCHIOL, DEVENDRA K. RASTOGI, JOACHIM WACHTER

Institut fiir Anorganische Chemie der Universitrit Regensburg, Universitcitsstr. 31, D-8400 Regensburg, F.R.G.

MARTIN MINELLI and JOHN H. ENEMARK*

Department of Chemistry, University of Arizona, Tucson, Ariz. 85 721, USA.

Received June 1.1984

In a short communication we reported that the diastereomers of la, differing only in their MO configuration, give two well separated and relatively sharp signals in the 95Mo NMR spectrum [l]. This result raised the question as to whether other diastereomers could also be distinguished by 95Mo NMR.

Therefore a general study of the 95Mo NMR spectra of four different types of compounds was under- taken.

All the compounds la-l, 2,3a, b, and 4a-c used for the study are depicted in the Scheme. For each compound that contains an optically pure ligand with (S)configuration at the asymmetric carbon atom C*, there are two diastereomers, Rr,.&o and SwOSc, differing only in the MO configuration. If a racemic ligand with (R)/(S) -configuration at the asymmetric carbon atom C* is used in the synthesis then there are four isomers, two diastereomeric pairs of enantiomers R,,So/S~,Ro and Srvr,S,-JR~,,Ro. In both cases, however, the same solution NMR spectra are obtained, which arise either from the two diastereomers or from the two enantiomeric pairs of diastereomers [2]. There are examples for both cases in the compounds of the Scheme.

Table I summarizes the ‘H and 95Mo NMR parameters of all the compounds, la-l, 2, 3a, b, and 4a-c. Column 3 of Table I gives the diastereomer ratio for each complex, determined by integration of appropriate ‘H NMR signals (usually C5H5). Columns 4-7 contain the chemical shifts and the chemical shift difference (A) of ‘H NMR signals, and columns 8-10 present the corresponding data for the “MO NMR spectra. For each diastereomer ratio differing from SO:50 the parameters for the excess diastereomer are italicized.

Column 11 in Table I gives the references for the synthesis and characterization of the compounds incorporated in the present study. The correlation of

*Authors to whom correspondence should be addressed.

the ‘H NMR and 95Mo NMR parameters of individual diastereomers with solubility, chromatographic behavior and absolute metal configurations (where known) can be found in the references.

The chelate ligands in compounds la-i are derived from 2-pyridinecarbaldehyde (R” = H) and primary amines (R = methyl, ethyl, isopropyl, benzyl; R’ = ethyl, phenyl, methyl ester, benzyl ester) [3, 41.

For these compounds the two diastereomers can, in all cases, be differentiated by ‘H NMR spectroscopy

[4-71. The chemical shift differences for the C5H5 resonances are large (0.3-0.4 ppm) if there is a phenyl substituent at the asymmetric center C* (nos.

1-3, Table I). However, two alkyl groups at the asymmetric center, as in compound Id, usually lead to C5H5 signals which are not resolved for the two diastereomers (no. 4). Sometimes, as in the case of Id, other signals can be used to determine the diastereomer composition. If an ester substituent is bonded to the asymmetric center C* in addition to a hydrogen atom and an alkyl group then the chemical shift differences for the CsH, resonances are around 0.1 ppm (nos. 5-7). The differences increase to about 0.4 ppm if there are benzyl substituents directly attached to the asymmetric center C* or in the ester group R’ (nos. 8,9). In addition to the chemical shifts for the CsH, resonances and their differences (columns 4 and 5, Table I), the same parameters for other ‘H resonances are given in columns 6 and 7, if appropriate.

The corresponding 9sMo parameters for compounds la-i are listed in columns 8-10 of Table I (nos. l-9). Similar to the ‘H NMR signals, the 95Mo NMR signals are well resolved for the diastereomers of la-c which have a phenyl ring at the asymmetric center C* (diastereomer splitting between 14 and 22 ppm), and the 9sMo signals are not resolved for the diastereomers of Id with two alkyl groups at the asymmetric carbon atom C*.

Isomer enrichment shows that the diastereomers of la-c with high field (shielded) CSH, signals also exhibit the high field (shielded) MO signals. The high field shift of the C5H5 signals in one of the diastereomers of compounds of the type la-c has been attributed to a conformation around the N-C* bond in which the phenyl ring closely approaches the CSH, ring [4]. The weak attraction arising from this C6Hs/C5H5 interaction has been called the fi-phenyl effect*. It is responsible for the high shift of the CsHs signal and probably also of the MO signal because in this conformation CsHs and MO are in the inner anisotropy region of the phenyl ring at C*.

*For a recent review concerning the p-phenyl effect see reference 8.

0020-1693/85/$3.30 0 Elsevier Sequoia/Printed in Switzerland

(2)

L6 Inorganica Chimica Acta Letters

H

“pC6H5

H3C

a b ^c ^d e t ₉ h I j

R CH3 C2H5 i-C3H7 CH3 CH3 QH5 i-CJH7 CH2w5 Cl-&H5 t ^CH3

R’ C6H5 C+H5 CgH5 C2H5 COOCH3 COOCH, COOCH3 CCOCH3 COOU$C$l, C~HI

R” H H H H H H H H H ICH3

o;pyp

I R

,,o”

c \

H/

^C6H5

H3C

1

a b

-I-+

^R ^H ^CH3

Scheme

I

4 ^C6H5

-

In contrast to ‘H NMR spectroscopy there is no diastereomer splitting in the 95Mo spectra if an ester group R’ is bonded to the asymmetric center C” (nos.

S-9). Even if there is an additional benzyl substituent at C* or in the ester group R’no 95Mo NMR splitting is observed, whereas such changes cause additional signal separation in the ‘H NMR spectrum (nos. 8,9).

Derivatives lj-1 carry methyl or phenyl substituents instead of H at the imine carbon atom of the five-membered ring. It has been argued that substitu- tion of H by CHa and C6H5 at the imine carbon atom of the chelate ring changes drastically the conformation of the C*H(CH,)(C,H,) group with respect to the MO fragment [9]. This is obvious from the C5H5 signals of the two diastereomers of lj which are almost isochronous and from the 95Mo signals, which are separated by 6 ppm (no. 10) compared to the 14 ppm of la. For compound lj the diastereomer

splitting is much more distinct in the 95Mo signals (6 ppm) compared to the C5H5 signals, but the diastereomer ratio is best determined by integration of the CHCHa signals, which are nicely separated.

The 95Mo diastereomer splittings of compounds lj-1 (R” = CHa, C6H5) are completely analogous to their counterparts la-i, with R’ = H. Compounds lj and 11 have a phenyl substituent at C* and show diastereomer splitting, whereas lk has only alkyl substituents at C* and does not show diastereomer splitting (nos. 10-12).

In contrast to the cationic complexes 1, compounds 2 and 3 are neutral. As chelate ligands they contain an o-metalated benzaldimine (2) [lo] and the anion of the Schiff base derived from (S)(-)-l- phenylethylamine and 2-pyrrolecarbaldehyde (3a)

[ 1 I] or 2-acetylpyrrole [3b] [ 121. All three compounds (2, 3a, b) contain a hydrogen/alkyl/phenyl

(3)

TABLE I. Diastereomers of Compounds la-l, 2, 3a, b, and 4a-c (acetone solution): ‘H NMR Spectra (6, i-TM& Bruker WM 250) and g5Mo NMR spectra (S,2M, NasMo04 in 2 HrO, pH 11, Bruker WM 250).

B % ; 8’

No. Compound ratio Cd5 A CH3 A “MO A Line Ref. 0 6 [wml [wml 6 bpml [wml 6 [wml [wml width [Hz]

z 3 %

1 la 25~75 6.03;5.66 0.37 _ _ -154; -168 14 <lOO, <lOO 1,4-7 2 lb 25~75 6.00; 5.64 0.36 - _ -145; -167 22 150,70 4 ; 3 IC 30:70 5.94 ; 5.55 0.39 - - -143; -166 20 100,70 4 : 4 Id 75:25 6.05; 6.05 l.62;1.57a 0.05 -186 120 4 2 - - 5 le 50:50 6.10; 6.03 0.07 ;i: _ - -169 - 90 4 a 6 If 60:40 6.09; 6.00 0.09 3.89;3.8Jb 0.04 -159 _ 90 4 7 lg 40:60 6.09; 5.95 0.14 3.93; 3.85b 0.08 -145 - 90 4 8 lh 30:70 6.01; 5.56 0.45 3.85; 3.78b 0.07 -160 _ 120 4 9 li 55:45 5.92; 5.58 0.34 _ - -156 - 160 4 10 lj 50:50 6.02; 5.99 0.03 2.11; 1.80 0.31 -141; -147 6 go,90 9 11 lk 40:60 6.05; 6.04 0.01 _ - -154 _ 90 9 12 11 75~25 6.03;5.30 0.72 _ _ -128; -160 32 130,180 9 13 2 50:50 5.25;4.87c 0.38 _ - -383; -396 13 50,150 10 14 3a 50:50 5.52;5.24 0.28 - - -293; -310 17 70,70 11 15 3b 50:50 5.57; 5.25 0.32 2.29; 2.16d 0.13 -293; -299 6 90,120 12 16 4a 50:50 _ _ 1.40; 1.37a*= 0.03 -339; -354 15 170,220 13 17 4b 50:50 _ - 1.20; 1.07a,c 0.13 -355; -370 15 160,160 13 18 4c 50:50 2.20; 2.16e 0.04 1.82; 1.76a 0.06 -298; -309 11 200,200 13 ‘CHCHs doublets. bCOOCH3 singlets. c In CDCls . dN=C(CH3). ePNCH3 doublets.

(4)

L8

combination at the asymmetric carbon atom C* and show clear 9sMo splitting of the diastereomers (nos.

13-15), similar to compounds la-c, lj, and 11.

Compounds 4 are polypyrazolylborato derivatives.

4a is a trispyrazolylborate complex; 4b and 4c are tetrapyrazolylborate complexes. 4a is methyl sub- stituted in the 3,5-positions [ 131. The optically active ligands are the aminophosphines (CeHs)sPN- (R)CH(CH,)(C,HS) with R = H (4a, b) and R = CHa (4~). The ligands CO and NO complete the coordina- tion shell of the MO atom. Although the chirality at the MO atom is only due to the differences of the CO and NO groups, the diastereomer splitting is re- markably high (nos. 16-18). This is in accord with other results showing that the structurally similar linear diatomic ligands CO and NO are stereochemi- tally distinctly different [ 141.

The 95Mo signals for the four different types of compounds appear in different chemical shift regions:

la-l between -130 and -185; 2 around -390; 3a, b between -290 and -310; and 4a-c between -300 and -370 ppm. All of these regions are in the deshielded part of the known chemical shift range for Mo(I1) monomers [ 1.5, 181.

The diastereomer ratios determined from integration of the 95Mo spectra are in accord with the ratios obtained from integration of the ‘H NMR spectra.

For rapid differentiation of diastereomers ‘H NMR is superior to 95Mo NMR because of the greater sen- sitivity and narrower lines in ‘H NMR. The spectral differences for two typical examples are shown in Fig. 1.

A--I

6 5.5 300 Loo PPm

Fig. 1. Left: CsHs region of the ‘H NMR spectrum of Ic in acetone solution (Bruker WM 250). Right: 9sMo resonances of 4b, 2 M Na2Mo04 in H20, pH 11 (Bruker WM 250).

Inorganica Chimica Acta Letters

One advantage of heteronuclear NMR studies of diastereomeric complexes is that such studies provide direct information about the transfer of chiral information among different parts of the molecule. ‘H NMR by itself provides information about the transfer of chiral information to other peripheral groups which contain protons. 95Mo NMR of diastereomeric molybdenum complexes directly probes the transfer of chiral information from the ligands to the metal center. In some cases, e.g. le-i, diastereomers cannot be detected at the metal center by 95Mo NMR even though relatively large differences are seen in the ‘H NMR of the C5H5 rings. In other cases, e.g. lj, sizable 95Mo splittings are observed, but the ‘H NMR of the CsH, rings are nearly isochronous for the two diastereomers.

For the present study, the wealth of ‘H NMR data for diastereomeric molybdenum complexes provided the impetus to explore 95Mo NMR as a direct probe of diastereomeric metal centers. However, for MoOz- (LcYsOR)~ complexes, direct observation of diastereomers in the 95Mo NMR spectra prompted reexamination of the ‘H and i3C spectra for evidence of diastereomers [ 191.

Acknowledgements

We thank the United States Department of Agri- culture (Grant No. 59-2041-l-626 to JHE) for support of this work. We thank Dr. Kenner Christen- sen for assistance with the 95Mo NMR measurements.

We thank the Deutsche Forschungsgemeinschaft, the Fonds der Chemischen Industrie, and the BASF ^AG for support of this work.

References 1 2 3 4 5 6 7 8 9 10 11 12

M. Minelli, T. W. Rockway, I. H. Enemark, H. Brunner and M. Muschiol, J. Organomet. Chem., 217, C34 (1981).

H. Brunner, I. Bauer and R. Lukas, 2. Naturforsch., Teil B:, 34, 1418 (1979).

H. Brunner and W. A. Herrmann, Chem. Ber., 105, 770 (1972).

H. Brunner and D. K. Rastogi, Inorg. Chem., 19, 891 (1980).

H. Brunner and W. A. Herrmann, Chem. Ber., 10.5, 3600 (1972).

H. Brunner and W. A. Herrmann, Chem. Ber., 106, 632 (1973).

I. Bernal, S. J. LaPlaca, J. Korp, H. Brunner and W. A.

Herrmann, Inorg. Chem., 17, 382 (1978).

H. Brunner, Angew. Chem., Int. Ed. Engl., 22, 897 (1983).

H. Brunner and D. K. Rastogi, Bull. Sot. Chim. Belg., 89, 883 (1980).

H. Brunner and J. Wachter, J. Organomet. Chem., 107, 307 (1976).

H. Brunner and W. A. Herrmann, J. Organomet. Chem., 63, 339 (1973).

H. Brunner and P. Beier, unpublished results.

(5)

horganica Chimica Acta Letters L9

13 E. Frauendorfer and H. Brunner, J. Organomet. Chem., 240, 371 (1982).

14 H. Brunner, Ann. N. Y. Acad. Sci., 239, 213 (1974).

15 S. Dysart, I. Georgii and B. E. Mann, J. Organomet.

Chem., 213, Cl0 (1981).

16 J. Y. LeGall, M. M. Kubicki and F. Y. Petillon, J. Organo- met. Chem., 221, 287 (1981).

17 R. T. C. Brownlee, A. F Masters, M. J. O’Connor, A. G.

Wedd, H. A. Kimlin and J. D. Cotton, Org. Magn. Reson., 20, 73 (1982).

18 M. Mine& A. BeIl, J. H. Enemark and R. A. Walton, Inorg. Chem., submitted for publication.

19 I. Buchanan, M. Mine& M. T. Ashby, T. J. King, J. H.

Enemark and C. D. Garner,Znorg. Chem, 23,495 (1984).